Katalytische valorisatie van lignine in de bioraffinaderij

Lignocellulose, the most abundant biomass constituent, represents a promising renewable resource for the production of biofuels, chemicals and materials. In analogy with a petrochemical refinery, lignocellulose can be processed in a so-called biorefinery. While a large number of biorefinery strategies have been described and investigated, new ones are constantly being developed. Most of these methods enable an efficient valorization of the carbohydrates cellulose and hemicellulose, but lignin is mostly obtained as a low-value, degraded residue or precipitate, that is generally used as energy source to fuel the process. The fact that lignin is the largest renewable source of aromatics and that lignin valorization is regarded as essential for the economics of the biorefinery stimulates a higher-value use of this substrate. In this regard, the conversion of lignin to chemicals presents a very promising valorization route. However the lignin streams generated in most biorefinery schemes are not suited as feedstock for chemicals. Therefore, valorization of lignin into chemicals should be a focal point of future biorefinery research. This doctoral research presents a lignocellulose biorefinery method that enables an efficient valorization of lignin into phenolic compounds. In this process, denoted as reductive lignocellulose fractionation, lignocellulose is treated in liquid phase under mild reducing atmosphere in the presence of a redox catalyst. The process involves the extraction and depolymerization of lignin from the lignocellulose matrix through combined solvolytic and hydrogenolytic action, yielding a lignin oil rich in phenolic monomers, next to di- and oligomers, and a solid carbohydrate pulp. The carbohydrate pulp can be further processed into chemicals or used for paper or pulp production. In order to efficiently separate the carbohydrate and lignin product families, both a high lignin removal from the pulp (delignification) and high carbohydrate retention in the pulp are important. The choice of solvent is found to significantly impact both the pulp retention and delignification efficiency. By testing a range of bio-derivable solvents with varying properties in the Pd on carbon-catalyzed reductive fractionation of birch wood, it is shown that a high solvent polarity is beneficial for delignification, but that a too polar solvent, like water, gives rise to significant carbohydrate solubilization. In order to express the efficiency of the process as a combination of delignification efficiency and pulp retention, a new empirical descriptor was introduced, denoted as ‘lignin-first delignification efficiency’ (LFDE). Methanol and ethylene glycol reached the highest LFDE values and were therefore indicated as most suitable solvents for the reductive fractionation. Processing in both solvents exhibits substantial differences in process and product characteristics, which are discussed extensively. Reductive lignocellulose fractionation enables a high yield production of phenolic monomers, which can be further upgraded to various chemicals. The monomers contain ortho-methoxy groups, next to various para-substituents, and removal of the methoxy groups (demethoxylation) is an important step in the upgrading of the phenolic monomers. In this work, the demethoxylation of a class of phenolic monomers, namely substituted guaiacols, is investigated with nickel on oxide catalysts. Two upgrading routes are studied, namely (i) demethoxylation towards substituted phenols and (ii) demethoxylation combined with aromatic ring hydrogenation towards substituted cyclohexanols. It is shown that in the conversion of guaiacol, most nickel on oxide catalysts show aromatic ring hydrogenation activity, but only Ni on CeO2 and ZrO2 allow a high yield production of cyclohexanol. The conversion of guaiacol to cyclohexanol is demonstrated to proceed through two pathways, with 2-methoxycyclohexanol and phenol as respective intermediates. Transformation of 2-methoxycyclohexanol is the rate determining step and requires a high reaction temperature. For Ni on CeO2, the catalyst activity could be effectively increased by increasing the nickel loading. Remarkably, an increase in average nickel particle size, resulting from an increasing nickel loading, was found to enhance the activity of the accessible Ni sites, which is indicative of structure-sensitive nickel catalysis. Ni on CeO2 enables a selective conversion of various substituted guaiacols, even of real lignin-derived 4-n-propylguaiacol, to the corresponding alkylcyclohexanols. To illustrate a possible valorization of alkylcyclohexanols, the stepwise conversion to alkylcaprolactone was demonstrated, involving dehydrogenation and Baeyer-Villiger oxidation. Alkylcaprolactones can constitute novel building blocks for bio-polyesters. Selective alkylphenol formation was only possible with Ni on a specific TiO2 support, namely anatase TiO2. Catalysts comprising a rutile or a mixed rutile:anatase TiO2 support exhibited a completely different catalytic behavior, pointing to an important role of the crystal phase in the catalytic transformation. Next to the support crystal phase, the nickel loading was also found to significantly impact the catalytic behavior and only for a well-defined range of nickel loadings, selective alkylphenol formation was observed. This range was denoted as critical loading range. When the nickel loading goes beyond the critical loading range, the product selectivity shifts towards saturated products, i.e. products with a hydrogenated aromatic ring, which implies a change in the nickel phase. Furthermore, the rate of alkylphenol formation was shown to increase with increasing specific surface area of the catalyst, but not with increasing nickel loading within the critical loading range, suggesting that the active sites for the demethoxylation are located on the TiO2 surface. These sites are proposed to be oxygen vacancy sites generated by hydrogen spillover from the nickel particles. The most remarkable finding was that demethoxylation of 4-alkylguaiacols over the Ni on TiO2 catalyst yields mainly 3-alkylphenol, instead of the expected 4-alkylphenol. It is shown that the catalyst exhibits the extraordinary activity to isomerize 4-alkylguaiacols to 5-alkylguaiacols, and convert the latter to 3-alkylphenols via demethoxylation. 3-Alkylphenols like 3-ethylphenol and 3-n-propylphenol are of special interest since they can be used as odor baits for tsetse flies, which constitute the main transmitter of the African sleeping sickness.nrpages: 175status: publishe